Electing Differently

Until June 30, 2011, the federal parliament has to pass a new version of the electoral law in Germany. This version is to exclude negative weights of voting that can come into being according to the electoral law in force. Searching for a solution, a relatively simple, but consequential and clear electoral system is presented: The Personalized Proportional System (PPS). This electoral system, which has its origins in a combination of the Finnish and the current German system, contains a clear division of roles: While the parties do nominate their candidates, the voters decide which of the candidates enter the parliament. There is a fixed number of parliamentary representatives (x, in Germany 598) that are elected in a fixed number of equal election wards (x/2, in Germany 299). Each voter has one vote to elect both, one candidate and his or her party. The numbers of seats that the parties get correspond with their nation-wide portions of votes, provided that they have got 5% or more of all valid votes. Successful candidates are: a) the candidates which have got the most votes in their election wards, b) the candidates which got the highest portions of votes, compared with the other candidates of their parties. This election system not only complies with demands of proportionality and consistency; it also motivates candidates and voters to actively participate in polls.

According to the current German law of elections, too many votes for one party may, under some conditions, reduce the number of representatives won by this party (negative weight of voting). The articles of the electoral law making this paradoxical phenomenon possible have been pronounced as unconstitutional by the federal constitutional court of Germany in July 2008.[1] The parliament has to pass a correspondingly changed version of the law until June 30, 2011.

A proposal for solving the problem goes to immediately clear sub-national surplus seats of a party in the nation-wide distributing process of party seats in parliament: If a party gets more sub-national seats by direct personal election than by votes, it receives correspondingly fewer seats in other sub-national states. Consequently, the maximal number of 598 deputies in German parliament is be kept strictly.[2] By this procedure there is no longer any risk of developing negative voting weights and no surplus seats come into being at national level. However: In case the Christian Social Union (CSU) wins surplus seats in Bavaria, these seats are kept at the national level (because the CSU is the only regional party which is able to overcome the 5% hurdle of nationally processing votes) – a distortion. And between different sub-national lists of a party, a disproportion develops because surplus seats, cast by relative majority, in one sub-national state restitute party seats, cast by proportional election, in another sub-national state.[3]

The search for a solution of the problem, indeed, could be perceived as a chance. The best way to exclude inconsistent rules is to develop a clear electoral system that complies with basic needs of the voters. So also other contradictions and defects of the current electoral system could be excluded and a voter friendly electoral right could be established. In the following, such an alternative is to display, the Personalized Proportional System.

Electoral systems deal with what the voters decide about, how they can take a vote and how votes are transformed into parliamentary seats. These regulations can be evaluated according to different criteria, such as structures, functions and special interests of involved actors.

Structurally, majority vote systems and proportional systems are facing each other: While proportional systems are about transforming votes as precisely as possible into seats, majority vote systems point at building stable governmental majorities by using disproportional mechanisms. By this way, two different electoral functions are emphasized, proportional representation and political stability (concentration). Additional electoral functions are the participation of the electorate, consistency and easiness to manage of an electoral system, along with institutional acceptability. In order to evaluate electoral systems in a realistic way, also special interests of involved actors, such as interests of certain parties or powerful persons, should be taken into consideration.

There are some interrelations between these criteria. So there is a transfer relationship between representation and stability (concentration). Also participation und easiness to manage often reduce each other. Governmental parties try to stabilize their governmental power by majority vote systems, while small parties usually are interested in getting proportional systems. Amongst the OECD countries, proportional systems have got more and more influence during the last decades. Even Great Britain, the mother country of the majority vote system, has discussed electoral reforms towards a more proportional structure.[4] And also in India, the biggest democracy of the world that by now operates with a modified majority vote system, some actors try to push a transition to a proportional system.[5]

According to the regulation in force, the members of the German federal parliament are elected in a specifically mixed electoral structure (Mixed Member Proportional System).[6] Out of 598 deputies, 299 deputies are personally elected by majority rules in election wards, while the other (299) deputies are determined through formerly established lists of the parties. The more votes a party gets, the more candidates established at the list of this party enter the parliament. If a party gets more seats by personal election than it is entitled to get through its portion of party votes, it keeps these surplus seats without any compensation for the other parties. That is: The more surplus seats in the parliament, the more the proportional system is watered down by majority vote elements. How many surplus seats for a party come into being at national level is calculated between the portions of votes and the surplus seats a party gets in different sub-national states – a very complex and often contradictory process of calculation. One arising contradiction is the developing of negative weights of voting.

A long time this electoral system has been considered as a factor of stability because through decades governments could be formed without big problems. In all cases, governmental coalitions came into being that based on clear political majorities. And there was a high turnout during the first decades of the new republic, particularly in the 1970s (up to 90% and more). Since the German reunion, however, the situation changed: It became more difficult to build stable governmental coalitions because the post-communist left regularly got parliament seats, but the other parties did not consider it as an acceptable coalition partner. Finally, the election turnout clearly dropped (down to 70,8% in the last German election 2009 and down to 43,3% in the last European election 2009:).[7]

That the current German electoral system cannot any longer be associated with uncomplicated governmental formation and high turn outs may seem not very hard to get over. Not acceptable, however, are grave defects of representation which characterize this electoral system. One of these defects is the sketched negative weights of voting: By logical contradictions between different regulative elements of the calculation of surplus seats of different sub-national states, a party may loose seats in case this party gets too many votes…This non-sense was object of the public attention during some weeks in case of a by-election 2005 in Dresden.[8] It is, indeed, no single case, but may appear any time in the framework of the current over-complex and contradictory system.[9]

Also apart from the issue of negative voting weights, surplus mandates mean different success values of votes. Because a vote for a party that does not get surplus seats has a lower value than a vote for a party that reaches surplus seats. And surplus seats come into being through majority election, that is, through the loss of all votes that are not successful. This double electoral disproportion appeared as long as a marginal problem as some parties regularly received very high shares of party votes. With the meanwhile developed five-party-system, the times are changing: The Christian parties and the Social democrats, it’s true, keep winning big portions of personal seats; their share of party votes, however, has clearly dropped – CDU/CSU to 33,8%, SPD to 23, 0% in the last national election 2009. Following, surplus seats come regularly into being. This holds in an extent that can decide about which governmental coalition comes to power – a fundamental problem of legitimation in a system that, all in all, should be a proportional system.

Different success values of votes, after all, come into existence through the 5% hurdle of processing votes into seats and the so-called Grundmandatsklausel that is to qualify the 5% hurdle (by giving access to parliament for the whole fraction with only 3 personal seats). In evaluating these two restrictions of proportionality, the 5% hurdle seems to be no significant distortion. Because this electoral institution fulfills a fundamental stability function and holds generally for any involved actor in the same way. Contrasting with this element of regulation, there is a clear misrepresentation through the so-called Grundmandatsklausel: By example, a party that gets 4,9% of all votes does not enter the parliament, while a party getting 3 % and 3 personal seats enters the parliament with 18 representatives, a sharp distortion.

Unequal values of votes, finally, do arise according to the current German electoral law through accepted deviations of the extent (numbers of voters) of wards: This extent is allowed to regularly vary by 15%. Even deviations up to 25% may be accepted according to the law in force.[10] Votes in certain wards, for example wards in rural areas with a lower density of population, can be more worth than votes in wards in areas with higher density of population.

In the given electoral system, voters have, on the one hand, the choice between different personal candidates of different parties, on the other hand, the choice between different parties. However, the voter cannot choose a certain candidate of his favored party. In this construction, a direct (personal) candidate who has not been elected by the voters, can neverthefewer enter the parliament through sitting on a sufficiently good place in the sub-national list of his/her party – an option that apparently contravenes against basic democratic principles. Facing this fact, it is a kind of lie to call the current electoral law in Germany a personalized proportional system.

The result of evaluating the current electoral system in Germany according to the criterion of clarity is unambiguous: Already the differences between the first vote (personal vote) and the second vote (party vote) are not clear for many citizens. Accordingly there has developed a kind of routine of newspapers to try to make the different votes understandable in the pre-phase of elections. Only few citizens are familiar with more specific elements of the complex German electoral law, such as surplus seats and “Grundmandate”. And details of the calculation process in relationship between party lists and surplus mandates are not understandable even for well-skilled observers. All in all, the German electoral system in force must be evaluated as very unclear, not friendly for voters.

The current electoral system in Germany is no clear proportional system and the voters are not entitled to choose candidates of their favored party personally. The system is contradictory and not clear (transparent). That’s why alternatives to the system should be thoroughly discussed.

I see two relevant alternatives to the current electoral system in Germany, the immediate clearance of surplus seats and the Personalized Proportional System.

For the elimination of negative weights of voting, the mathematician Friedrich Pukelsheim has developed a solution which sometimes is called elegant, the immediate settlement of surplus seats.[11] According to the German law, the number of parliamentary seats that a party is entitled to get as result of an election is determined through a divisor without taking into calculation the number of surplus seats. Through this separation, negative weights of voting are possible. Facing this problem, Pukelsheim proposes to combine the nation-wide regulation of divisors and the regulation of sub-national surplus seats. He calls the resulting concept direktmandatsbedingte Divisormethode mit Standardrundung.

Following this concept, each sub-national list of a party gets so many seats as it is entitled to get by dividing the nation-wide sum of party votes through a rounded divisor, in the minimum the number of seats won by this party directly (by majority procedure). According to this formula, surplus seats at the sub-national level remain in any case: If a party gets more direct seats than it is entitled to get by their party votes, this sub-national surplus seats are taken into consideration in determining the nation-wide divisor. Because now fewer seats can be freely distributed, the divisor increases. Corresponding with this increase, other sub-national lists of the considered party get fewer seats. That is: Surplus seats of a party in certain sub-national states that have come into being in a majority procedure, are mathematically processed with seats that come into being by a proportional procedure, favoring the majority element of regulation.

By this procedure, the principal number of deputies in the parliament (598) is usually realized because no surplus seats emerge any longer at national level.[12] Any risk of negative weights of voting to emerge is eliminated. Indeed, if the Christian Social Union (CSU), a very strong regional party, wins surplus seats in Bavaria overcoming the nation-widely5% hurdle, these seats will remain – a distortion. Also between different sub-national lists of parties, a disproportion favoring the majority elements of the system develops: Surplus seats in one sub-national state then replace regularly won party seats in other sub-national states.

The second alternative, the Personalized Proportional System, has been developed in a combination of Finnish and German electoral rules. It equals the current electoral system in Germany in some respects: As until now, there are 598 representatives in parliament and 299 wards. Every party gets as many seats as it is entitled to get by its percentage of votes including a 5% hurdle for transforming votes into seats. Like the current procedure, any party nominates one candidate for one ward. The winner of a ward, who enters the parliament, is elected by a relative majority of votes.

Aside of these similarities, however, there are some changes: Every voter only disposes over one vote for electing both, one candidate and his or her party. Which ones of the candidates enter the parliament, depends on a) The shares of votes they receive, b) the number of party deputies entering the parliament. For example: Corresponding with its share of votes, a party sends 100 representatives to the parliament. 40 candidates of this party have won their wards by relative majorities. Amongst the rest, those enter the parliament who have received the 60 highest portions of votes in their wards.

According to this procedure, it is very unlikely that surplus seats come into being. Because for this, the number of (ward) winners must be higher than the number of all nation-widely (by party votes) elected representatives. In this unlikely case, the shares of votes that a party receives in wards without own winners must be extremely low. See the following example: A party sends, in accordance to nation-widely received 20% votes, 120 of 598 representatives to the parliament. The candidates of this party won 123 out of 299 wards, based on a average voting share of 40%. In this case, the party had to receive not more than 6.7% of the votes in the wards where their candidates did not win. By any more winner of a ward, this share must be even lower than 6.7%,

If surplus seats came into being, those seats should not be balanced by seats for other parties.

The wards should encompass equally many persons entitled to vote. For managing this, the numbers of entitled persons should be counted or estimated 18 months before each regular election. Not later than every 8 years all data should be empirically verified. Based on the currently existing data, at the latest 12 months before the election the electoral wards are fixed.

After an election, the votes are counted and the shares of the single candidates as well the shares the parties have received are determined. In doing this, the 5% hurdle of transforming votes into seats and the highly proportional rounding method of Saint-Lague have to be used.[13] A ranking list for any party shows the percentages of all candidates and the consequences for entering the parliament.

Different from the current system, the presented electoral system deserves the nomination Personalized Proportional Election. That’s why the system is, apart from the necessary 5% hurdle, a pure proportional system. And the voters not only elect parties but also personally every candidate.

The one-vote system is very easy to handle: Wrong interchanges between a first and a second vote are excluded and the numbers of votes as well as the numbers of seats for any party and candidate can be very easily counted respectively calculated. That’s why there are no sub-national surplus seats and sub-national party lists any longer. Rather, the election results can be nation-widely summed up and transformed into seats. Also the election results of single candidates can be read on a simple ranking list of the single parties.

At the end, the presented personalized proportional system fosters the motivation of poll. According to this electoral system, not only the winner of a ward has a chance to enter the parliament, but also other candidates. There are no party lists any longer that transport mighty candidates into the parliament without any legitimization by the people, and the people will be even more fascinated by the real electoral competition. All in all, this alternative is clearly structured; it complies with demands of a representative proportional election and implies a clear division of roles: While the parties do nominate their candidates, the voters decide which ones of the candidates enter the parliament.

The most people consider electoral regulations as simply fixed. From sheer force of habit, they do not question currently existing electoral law. Against this background, electoral systems usually belong to political cultures of nations. Accordingly, electoral systems have kept up for decades or even centuries in many countries, even if they contain contradictory or/and anachronistic elements. If fundamental reforms or processes of creation in this field came into being, they did it because of deep crises or after establishing new states.

Reforms of electoral systems nevertheless have been realized in some cases, even in competitive welfare-states.[14]

In those cases, often stimulating external ideas are of influence. Usually controversies develop to questions such as: Is it reasonable to change the existing electoral law? What about the political scenery of interests regarding a reform? What alternative does comply best with relevant demands and challenges of our country? How to make a reasonable proposal well-known and acceptable in politics?

--------------------------------------------------------

[1] § 7 Abs. 3 Satz 2 in Verbindung mit § 6 Abs. 4 und 5 Bundeswahlgesetz. Zur Pressemitteilung des BVG siehe: http://www.bverfg.de/pressemitteilungen/bvg08-068.html

[2] Vorschlag von Friedrich Pukelsheim: http://www.math.uni-augsburg.de/stochastik/pukelsheim/2008Berlin/VorschlagBWahlG.pdf, aufgegriffen von der Fraktion Bündnis 90/Die Grünen im März 2009 (Bundestagsdrucksache 16/885: http://dipbt.bundestag.de/dip21/btd/16/118/1611885.pdf, zur Stellungnahme von Volker Beck: http://www.gruene-bundestag.de/cms/archiv/dok/273/273708.aenderung_des_bundeswahlgesetzes.html

Die SPD stand dem Antrag positiv gegenüber, fühlte sich aber als Noch-Regierungspartei in der Pflicht gegenüber ihrem Regierungspartner CDU/CSU. Der Antrag verfehlte somit eine Mehrheit im Bundestag.

[3] Artikel 7a, Absatz 7 des Entwurfs von Bündnis 90/Die Grünen (Anmerkung 2).

[4] http://www.historylearningsite.co.uk/proportional_representation.htm

[5] http://www.scribd.com/doc/20355033/C-E-R-I-Introduction

[6] Artikel 1 Bundeswahlgesetz

[7] http://de.wikipedia.org/wiki/Wahlbeteiligung#Bundestagswahlen; http://de.wikipedia.org/wiki/Europawahl

[8] Nachwahl im Bezirk Dresden 1, 14 Tage nach der Bundestagswahl

[9] http://www.wahlrecht.de/news/2009/08.htm

[10] Artikel 3, Absatz 1, Punkt 3 Bundeswahlgesetz: http://www.gesetze-im-internet.de/bwahlg/__3.html

[11] http://www.math.uni-augsburg.de/stochastik/pukelsheim/2008Berlin/VorschlagBWahlG.pdf

[12] Vorschlag von Friedrich Pukelsheim: http://www.math.uni-augsburg.de/stochastik/pukelsheim/2008Berlin/VorschlagBWahlG.pdf, aufgegriffen von der Fraktion Bündnis 90/Die Grünen im März 2009 (Bundestagsdrucksache 16/885: http://dipbt.bundestag.de/dip21/btd/16/118/1611885.pdf.

[13] http://de.wikipedia.org/wiki/Sainte-Lagu%C3%AB-Verfahren

[14] Zum gesamtsystemischen Überblick für West- und Osteuropa siehe Wolfgang Ismayr (Hrsg.): Die politischen Systeme Westeuropas Wiesbaden 2009 (4. Aufl.) und: Die politischen Systeme Osteuropas 2009 (3. Aufl.).

[1] Artikel 7a, Absatz 7 des Entwurfs von Bündnis 90/Die Grünen (Anmerkung 2).

[1] Zum gesamtsystemischen Überblick für West- und Osteuropa siehe Wolfgang Ismayr (Hrsg.): Die politischen Systeme Westeuropas Wiesbaden 2009 (4. Aufl.) und: Die politischen Systeme Osteuropas 2009 (3. Aufl.).